Effects of Cloud on Land Surface Temperature (LST) Change in Thermal Infrared Remote Sensing Images: a Case Study of Landsat 8 Data

摘要

Clouds exert a great influence on land surface temperature (LST). It remains unclear how clouds affect the LST change in satellite images. The focus of the study is to examine the effects of cloud on LST change in Landsat 8 images over various ground surface types, cloud sizes and seasonal scales. The difference between solar and satellite geometry over the images provides the possible way to identify the impact of cloud in Landsat 8 images. The shadow of cloud represents the true effects of cloud on LST change. Using the available mono-window algorithm, we were able to retrieve the LST of the shadow for the analysis of the cloud effects. Four typical ground surfaces were selected for the analysis: desert/bare surface, vegetation, urban and water covered area. The LST difference between the shadow and its nearby clear-sky region represents the effects of cloud on LST change in the corresponding ground surface. The results indicated that the, highest LST difference was observed over dessert/barren land in the magnitude of 8.28°C. Moreover, when the cloud size relatively smaller or larger the effect on LST slightly varies. The summer season has the most remarkable effect of cloud on LST change. The water surface has the lowest LST difference in overall cloud analysis. These change of LST under cloud in various conditions can be helpful for the studies which are focus on LST estimation during cloudy conditions.